Discovery of crime event sequences with constricted spatio-temporal sequential patterns

Journal of Big Data

Table 7 Computation time (in sec) and the number of discovered patterns for the Pittsburgh Police Incident Blotter Dataset

R = 350 meters, T = 11520 (8 days)
\(PI_{min}\)	PI-strong ST seq. patterns						PI-strong closed ST seq. patterns		PI-strong CSTS patterns
	STS-Miner [9]		CSTPM [10]		STBFM [11]		CST-SPMiner [12]		CSTS-Miner \((\varepsilon = 0.025)\)		CSTS-Miner \((\varepsilon = 0.05)\)		CSTS-Miner \((\varepsilon = 0.075)\)
	Time	# Patterns	Time	# Patterns	Time	# Patterns	Time	# Patterns	Time	# Patterns	Time	# Patterns	Time	# Patterns
0.33	26	796	9	796	9	796	9	657	9	594	9	576	9	556
0.32	31	1002	9	1002	9	1002	9	819	9	712	9	681	9	666
0.31	39	1320	9	1320	9	1320	9	1065	9	883	9	827	9	802
0.3	56	1990	9	1990	9	1990	9	1573	9	1161	9	1050	9	997
0.29	116	4371	11	4371	10	4371	10	3323	10	2235	10	2018	10	1932
0.28	269	10,434	17	10,434	12	10,434	12	7741	12	4516	12	4206	12	4104
0.27	898	35,166	100	35,166	19	35,166	19	25,446	19	14,956	20	14,182	22	13,939
0.26	–	–	1658	143,666	47	143,666	47	100,235	51	60,977	67	58,907	99	58,519
0.25	–	–	–	–	206	1,073,917	206	683,860	253	515,128	712	509,232	1744	508,492